# True Fantasy Points: 2017 Quarterback Projections

A stats-based guide for separating skill from luck in quarterback projections.

Two years ago, I introduced my True Fantasy Points (TFP) system. For newcomers, my system is based on the fundamental truth that efficiency stats in football have a certain level of statistical noise (i.e., they're high-variance) and therefore uses math to properly regress players' Footballguys projections to the mean given their amount of projected opportunity (i.e., pass attempts, carries, and targets.). In the end, I use this calculation to compare players' "true," skill-based projected points with David Dodds' current projected points according to Footballguys standard scoring system. If you're still confused, click here and/or read the next paragraph.

Consider Eli Manning. In his Giants career, Manning's fantasy-relevant efficiency stats are 7.25 Yards per Attempt (YPA), 4.95% Touchdowns per Attempt (TD%), and 3.18 Interceptions per Attempt (INT%). Properly regress these values to the mean of all quarterbacks returning to their 2016 team, and Manning ends up with a 7.25 True YPA, a 4.91% True TD%, and a True 3.03% True INT%. Now, multiply these "true" efficiency stats to Dodds' current projection for pass attempts and apply the result to Footballguys' standard scoring system. The result: My TFP system projects 314.3 points for Manning, whereas Dodds projects 316.2 points, so he's currently "overrated" by 1.9 projected points.

# TRUE STATS THROUGH 2016

Below are actual passing stats and "true" passing stats for the 26 quarterbacks starting for the same team they played with last season (sorted by True YPA):

ACTUAL STATSTRUE STATS
PLAYERTMYPARKTD%RKINT%RKYPARKTD%RKINT%RK
Aaron Rodgers GB 8.01 1 6.56% 1 1.44% 3 7.95 1 6.17% 1 1.78% 2
Russell Wilson SEA 7.98 3 5.57% 3 1.97% 8 7.87 2 5.29% 4 2.23% 8
Ben Roethlisberger PIT 7.84 4 4.87% 12 2.48% 16 7.80 3 4.84% 11 2.51% 15
Philip Rivers LAC 7.72 6 5.21% 6 2.72% 22 7.69 4 5.12% 5 2.68% 23
Kirk Cousins WAS 7.78 5 4.63% 16 2.70% 21 7.68 5 4.65% 16 2.64% 21
Drew Brees NO 7.69 7 5.46% 5 2.40% 14 7.66 6 5.34% 3 2.44% 11
Dak Prescott DAL 7.99 2 5.01% 7 0.87% 1 7.66 7 4.78% 12 2.22% 6
Carson Palmer ARI 7.67 8 4.97% 9 2.59% 18 7.60 8 4.87% 9 2.59% 18
Tom Brady NE 7.59 10 5.68% 2 1.46% 4 7.57 9 5.51% 2 1.74% 1
Marcus Mariota TEN 7.61 9 5.48% 4 2.31% 12 7.50 10 5.03% 6 2.50% 14
Matt Ryan ATL 7.46 11 4.87% 11 2.17% 10 7.45 11 4.84% 10 2.28% 9
Cam Newton CAR 7.44 12 4.64% 14 2.66% 20 7.42 12 4.65% 14 2.63% 20
Tyrod Taylor BUF 7.42 13 4.53% 17 1.47% 5 7.37 13 4.62% 17 2.22% 7
Jameis Winston TB 7.37 14 4.53% 18 2.99% 24 7.34 14 4.61% 18 2.75% 24
Andy Dalton CIN 7.26 15 4.64% 15 2.65% 19 7.26 15 4.65% 15 2.62% 19
Eli Manning NYG 7.25 16 4.95% 10 3.18% 25 7.25 16 4.91% 7 3.03% 26
Andrew Luck IND 7.20 17 4.98% 8 2.57% 17 7.21 17 4.89% 8 2.57% 17
Matthew Stafford DET 7.15 18 4.37% 19 2.32% 13 7.16 18 4.44% 19 2.40% 10
Sam Bradford MIN 7.02 20 3.62% 24 0.91% 2 7.13 19 4.32% 21 2.17% 4
Trevor Siemian DEN 7.00 21 3.70% 23 2.06% 9 7.12 20 4.37% 20 2.47% 12
Alex Smith KC 7.04 19 3.99% 21 1.50% 6 7.08 21 4.23% 24 2.01% 3
Joe Flacco BAL 6.93 22 3.97% 22 2.43% 15 6.96 22 4.11% 25 2.47% 13
Blake Bortles JAX 6.59 23 4.04% 20 2.99% 23 6.72 23 4.29% 23 2.79% 25
Carson Wentz PHI 6.23 25 2.64% 25 2.31% 11 6.64 24 3.94% 26 2.51% 16
Derek Carr OAK 6.46 24 4.68% 13 1.79% 7 6.61 25 4.68% 13 2.18% 5
Jared Goff LAR 5.31 26 2.44% 26 3.41% 26 6.60 26 4.32% 22 2.67% 22

I spent the majority of this section in last year's article extolling the "true" statistical virtues of Aaron Rodgers and Russell Wilson as the two-most efficient quarterbacks in the NFL. Therefore, rather than doing a rehash, I'll instead extol the virtues of a hyperefficient newcomer to the scene. Namely, after regressing his 459 pass attempts to the mean, Dak Prescott appears to be the real deal. Of the 26 quarterbacks in the above table, only 6 rank in the top half for all three "true" stat categories: Rodgers, Wilson, Drew Brees, Tom Brady, Matt Ryan, and ... Prescott.

To put this in context and give some actionable fantasy football intelligence, here's where those six quarterbacks currently rank according to average draft position (ADP):

1. Rodgers is 16th.
3. Brees is 31st.
4. Ryan is 42nd.
5. Wilson is 55th.
6. Prescott is 85th.

Based on "true" stats, then, Prescott is poised to be a value. He's basically Wilson, just three rounds later; which is fitting because both have been perceived (incorrectly) in their careers as young passers riding the coattails of an elite running game. Yes, the Seahawks and Cowboys offenses have featured Marshawn Lynch and Ezekiel Elliott at running back in recent years, but Wilson and Prescott are (underappreciated) paragons of offensive efficiency in their own right.

What I'm trying to tell you here is that, if Elliott's suspension holds up, thereby necessitating an expanded passing role for Prescott, don't fret. In fact, double down. And if Elliott's suspension's delayed ad infinitum, also don't fret. (Prescott finished as QB6 last season.)

# TRUE PROJECTIONS FOR 2017

Alright, so multiplying the "true" per-attempt stat above by Dodds' projections for pass attempts, I get TFP, which we can then compare to Dodds' point projections. As a result, we arrive at the following table:

DODDS STATSTRUE STATSPOINTS
PLAYERTMYDSTDSINTSYDSTDSINTSDODDSTFPDIFF
Matt Ryan ATL 4509 32 11 4021 26 12 358.1 308.8 +49.2
Tom Brady NE 4639 36 8 4406 32 10 381.0 351.4 +29.5
Derek Carr OAK 4037 29 11 3803 27 13 322.4 300.7 +21.6
Andrew Luck IND 3802 28 15 3683 25 13 325.2 309.1 +16.1
Ben Roethlisberger PIT 4309 31 14 4334 27 14 328.8 313.6 +15.1
Jameis Winston TB 4223 27 15 4053 25 15 340.6 325.6 +14.9
Andy Dalton CIN 3906 24 11 3761 24 14 313.8 304.3 +9.5
Philip Rivers LAC 4320 31 14 4337 29 15 329.6 320.9 +8.7
Carson Wentz PHI 3713 22 13 3626 21 14 289.5 282.4 +7.1
Dak Prescott DAL 3725 24 10 3714 23 11 324.7 320.1 +4.6
Marcus Mariota TEN 3777 26 11 3786 25 13 329.8 326.2 +3.5
Eli Manning NYG 4243 29 14 4285 29 18 316.2 314.3 +1.8
Alex Smith KC 3337 20 9 3340 20 9 273.8 273.4 +0.4
Sam Bradford MIN 3280 20 10 3293 20 10 237.4 237.8 -0.4
Matthew Stafford DET 4002 23 11 3974 25 13 310.3 313.1 -2.8
Trevor Siemian DEN 3226 20 13 3254 20 11 242.0 245.1 -3.1
Drew Brees NO 4697 31 15 4672 33 15 353.1 358.3 -5.3
Russell Wilson SEA 3919 26 11 3975 27 11 348.2 353.5 -5.3
Cam Newton CAR 3623 24 12 3791 24 13 335.7 341.7 -6.1
Kirk Cousins WAS 4313 24 12 4347 26 15 325.1 333.1 -8.0
Blake Bortles JAX 3192 19 14 3224 21 13 262.7 271.3 -8.6
Tyrod Taylor BUF 3048 21 10 3282 21 10 300.5 310.5 -10.0
Joe Flacco BAL 3731 22 13 3906 23 14 272.9 285.0 -12.1
Jared Goff LAR 3250 19 15 3302 22 13 236.8 251.4 -14.6
Aaron Rodgers GB 4391 33 10 4609 36 10 393.2 414.9 -21.8
Carson Palmer ARI 3591 21 12 3825 24 13 255.9 280.5 -24.7

As an example of how to read the table, focus on the row representing Prescott. From left to right, what's displayed here is that Dodds projects him to throw for 3,725 yards, 24 touchdowns, and 10 interceptions, while Prescott's "true" stats imply that, over 485 projected attempts, he should throw for 3,714 yards, 23 touchdowns, and 11 interceptions. yards. Apply the Footballguys standard scoring system to these stats (also adding in rushing points), and it turns out that TFP and Dodds' point projection are in agreement; only 4.6 points apart.

Based on the above table, below are two quarterbacks whose "true" stats suggest they'll overtperform Dodds' projections, as well as two others likely to underperform.

## Underrated: Carson Palmer

Extrapolating from Dodds' projections, Palmer will post a 7.14 YPA in 2017, whereas his True YPA in Arizona is 7.60. Similarly, Dodds projects Palmer to have a TD% of 4.17%, whereas his True TD% is 4.87%. Given Palmer's advancing age (and Larry Fitzgerald's for that matter), as well as the continuing health concerns for John Brown, seeing a downturn in Palmer's viability as a fantasy football quarterback is reasonable. That said, seeing an efficiency downturn to the extent projected by Dodds is statistically unlikely.

## Underrated: Aaaron Rodgers

Rodgers' TFP is so much higher than Dodds' projection because 4,391 yards in 580 projected attempts translates to a YPA of 7.57, whereas we saw from the earlier table that Rodgers' True YPA in Green Bay is 7.95. It's worth noting how scary-good this makes Rodgers out to be. He's projected as QB1, and my "true" stats still suggest that's underestimating him. (Does QB0 exist?)

## Overrated: Matt Ryan

From both stats-based and tape-based perspectives, this summer has seen analysts proclaiming that Ryan is poised for a historic level of regression to the mean. When worlds collide like this, it's a good sign that it's true. And indeed, Ryan's True YPA of 7.45 dwarfs what Dodds projects (8.35), as does his True TD% of 4.84% (vs. Dodds' 5.93%). Couple this with the fact that Ryan's currently QB4 per ADP, and it's clear that he's TFP's No. 1 "avoid" candidate at the position.

## Overrated: Derek Carr

Longtime readers/followers will surely recognize that Carr's been a perennial member of my overrated list. As you might have noticed earlier in the first table, Carr ranks 25th of 26th in True YPA, just ahead of Jared Goff, who (arguably) may have had the worst passing season in the history of the NFL. Carr's "breakout" 7.03 YPA in 2016 barely moved his "true" YPA needle from 6.51 to 6.61. (Kudos to Dodds, however, for nailing Carr's 3,937 yardage projection on the number. An exact bid, if you will.) Heading into 2017, Dodds projects a YPA for Carr that's essentially the same as last season (7.02) rather than a regression to his "true" rate.

What's worse for Carr's fantasy football prospects this season as opposed to last is that Dodds also projects a TD% of 5.04% vs. his True TD% of 4.68%. And as was the case with Ryan, the problem presented by all of this math is exacerbated by Carr's current ADP of QB8.